On Decentralized Wireless (DeWi)

Decentralized Wireless (DeWi from here on in) is a burgeoning industry that aims to democratize wireless network infrastructures. Related to mesh networking, where multiple, spread-out routers/repeaters are located around an area to provide wireless network coverage (be it WiFi, LoRaWAN/IoT or 5G cell) instead of a single central router/tower. Cryptoeconomics are applied to DeWi models to bootstrap and incentivize network growth and coverage (i.e., tokens are used to solve the ‘cold start’ problem). In short, DeWi is the blending of mesh networking and cryptoeconomics that enables a novel approach to build wireless networks.

Helium, a pioneer of the space, start by providing LoRaWAN. Their LoRa network is specifically designed to power all things IoT. For context, legacy WiFi networks are too expensive and resource intensive (expensive hardware, small coverage radius, and built for large data transfers) as compared to LoRa if the goal was to solely power IoT-powered devices. They proved that building a LoRa network using cryptoeconomics works, boasting 15,000 hotspots on January 2021 and over 980,000 at the time of writing. However, traction from the users’ demand for the LoRa network has yet to show, and this was reflected in their revenue and thus the price action of their $HNT token. The user-side demand for 5G networks has been proven and is an immediate opportunity, unlike in LoRa where they entered a potential blue ocean. Realizing this, they pivot to building a network of 5G cells, with the broader DeWi sector skewing in the same direction.

For background, the global telco industry has always been centralized and highly capital-intensive. It is common around the globe to have 2–5 network providers powering a large majority of a country’s population. For the scope of this piece, let us focus more on the US telco industry.

The process for a network provider (e.g., Verizon) to build a new cellular network is as follows — 1) raise capital (tens of billions) through debt, 2) purchase spectrum licenses from the government, 3) contact manufacturers (e.g., ZTE, Huawei) to build required hardware, 4) partnering with tower providers (e.g., Crown Castle) for co-location of hardware and 5) regular maintenance of deployed devices. Unlike 2/3/4G networks, 5G networks use small cells, which have a per cell smaller coverage radius, thereby needing more cells/”routers” to provide similar coverage to legacy networks.

These five processes are highly CapEx and OpEx intensive, and DeWi plans to change this by offloading these costs to a distributed network of providers (i.e., miners) and using off-the-shelf equipment. 5G networks by design, require a large number of distributed cells, which is a perfect case of mesh networking. A clear benefit of using a mesh network is its ability to scale, be it strengthening the coverage of a dense area, or expanding to an area previously economically unviable to cover using central towers. All in all, DeWi results in superior unit economics than legacy systems.

How does it work (Helium, but can be broadly applied across DeWi networks)
Helium IoT network miners earn $IOT and 5G network miners earn $MOBILE. Both of these are redeemable to $HNT. This will be abstracted in the following segment.

Miners purchase off-the-shelf devices, provide internet and power to the device, and connect the device to the Helium network in less covered regions. Miners earn $HNT tokens for powering the network (transferring data) and securing the Helium blockchain. During the initial setup of their devices to the network, soon-to-be miners have to pay two fees, “add gateway” and “assert location”, which are both denominated in $HNT.

They utilize a work (validity) mechanism called Proof of Coverage, which enforces the trueness of routers’ location and uptime. Each router/enforcer can challenge another router’s trueness and will be slashed/rewarded depending on the outcome. All activities are recorded on the Helium blockchain, which runs on a Proof of Stake consensus mechanism.

For users to utilize the Helium network, they have to purchase $HNT tokens, which are to be burnt into data credits (DC), they pay “network fees” or rather data usage fees which are denominated in DC. The higher the utilization of the Helium network, the more $HNT tokens will be burnt.

DeWi powers a network by leveraging unlicensed spectrums (i.e., Citizens Broadband Radio Service (CBRS)), thus eliminating the need to purchase spectrum licenses. Due to the regulatory clarity on the unlicensed spectrum in the US and proven 5G demand, major efforts of 5G DeWi are exclusive to the US.

For perspective, telecom networks generated US$125 bn of EBITDA in 2021, and US$95 bn is paid out to the US government in the form of spectrum licenses and taxes.

As a result of 1) not having to pay spectrum costs, 2) not having to pay cell towers for co-location and 3) crowdsourcing both CapEx and OpEx, DeWi can deliver 5G at 50% of traditional telco’s cost basis.

Possible business models
For 5G DeWi networks, three main business models have been identified, neutral host, cryptocarrier, and private networks.

Neutral host
Neutral hosts provide the infrastructure (comparable to tower providers e.g., Crown Castle) for multiple mobile network operators (e.g., Verizon) to operate on.

Cryptocarrier
A crypto-native mobile network provider (comparable to Verizon, etc.) that serves end-users by leveraging its own network. The development of eSIM has enabled this business model to have a better edge than before as it significantly cuts down switching costs.

Private network
Private networks are built by renting out the existing DeWi network to institutions (e.g., IoT startups, airports, etc.) to provide bespoke network coverage (private, stronger, and reliable).

My choice
I believe that at Dewi’s current stage, neutral hosts show significant benefits over the other two.

Being a cryptocarrier means that there will be minimal intermediaries and we could minimize the fees to end-users. However, as a starting point of a new industry, directly jumping to the consumer side is tricky — 1) price insensitivity, are users willing to switch to a new carrier to save costs, and 2) crypto itself does not have the best public sentiment, how can we convince the end-users that this is sustainable.

Private networks are ideal. The anonymity, strength, and coverage can all be tweaked per the client’s requests. However, the market for private networks itself (US$1.4 bn) will not be enough to sustain a full-fledged DeWi network. Perhaps shortly, this model would be more sought after.

On the other hand, neutral hosts are a perfect balance of market demand and timing. Consumers may not be price-sensitive enough to shift over their network providers, but these mobile network operators are. As highlighted by Escape Velocity’s Sal, the unit economics of conventional mobile network operators are not the best, as a DeWi network, we have the upper hand between deals. By offloading their network activity into DeWi networks, they will be able to serve more customers better at a flexible cost, and DeWi miners can sustainably earn rewards. Of course, the greatest hindrance to this model is regulatory uncertainty. Given the fact that mobile network operators have a strong relationship with the government (spectrum licenses, real estate, towers, etc.), how will the government react to this move?

To truly empower miners, the best solution would be to pursue becoming a neutral host, whilst providing private network services.

Market landscape
As previously mentioned, the DeWi space is burgeoning. DeWi networks are increasingly shifting towards 5G. Hardware manufacturers produce the required device/router to participate in DeWi networks. Enterprise deployers are institutions that deploy devices at a large scale. Third-party distributors are retailers of hardware manufacturers, which may offer additional timely offers. Service providers/tools/marketplaces are community-built tools to assist miners/users navigate and maximize the Helium ecosystem.

Problems and thoughts
Helium’s LoRaWAN focus proves to be successful in building a network, but not in a business case. Figure 4 shows that the demand for Helium (data credits) has been significant, but looking deeper into the chart, a negligible amount comes from actual data usage (data transfer). A large majority of the revenue comes from hotspot setups (add gateways / assert location) and maintenance. This reflects the fact that most miners’ revenue comes from $HNT inflation and new miners’ setup, with no actual value accruing. It is debatable whether this model is viable.

However, my take on this is that it is not sustainable. They grew the network coverage too fast relative to user demand. It gives a huge parallel to Filecoin, a distributed storage network, where the supply side was growing extremely quickly relative to the user side thus diminishing the returns of the supply side. With no clear trajectory on the user side, this model will not work. Of course, the community, and the team acknowledges this, hence the pivot to 5G provision. Something to note is that, in Filecoin’s case, storage providers could simply use their PCs to “mine” the network. Whereas in Helium, miners have to purchase a dedicated device, thereby “forcing” miners to keep their devices on as long as their minuscule sustainable revenue could pay off the internet and power costs, in hopes to contribute to their upfront purchase of the device. However, this also means that a huge majority of miners are looking to sell to realize future profits, and as in the case of Bitcoin, miners’ selling is not a good indicator, they are the backbone of the network.

Not pointing fingers by any means, but two parties significantly benefit from this model. Device providers get massive amounts of sales, and early miners got a huge payout (high revenues and $HNT price) from the entrance of new miners.

Having a dedicated network for IoT itself is definitely not (yet) user-friendly, and from the user side, they are not engrained enough and thus not costly enough for them to seek alternatives from their conventional WiFi networks. Maybe in the future, when smart homes are considered homes, and smart devices are considered devices, LoRa’s benefits can be highlighted. From an institutional side (e.g., electric vehicle manufacturers, etc.), although the LoRa network shows significant cost cuts from the conventional cell network, this cost has always been passed on to end-users anyways. The end-users are not price-sensitive enough to demand this change, and cell networks currently still have better coverage than LoRa networks. For now, to at least maintain the network of miners, Helium/LoRa should begin by ensuring that manufacturers are shifting to dual connectivity (cell and LoRa support).

Cycling back, the main problem with their LoRa network is scaling supply-side too fast. If the IoT/LoRa network is indeed a blue ocean (figure 5), just how long can existing miners wait?

Following FutureMoney’s X2Earn paradigm, the broader DeWi sector falls under category 7 — low frequency, high capital, low work — and Helium specifically is in its phase of consolidation (user discovery). Helium is in a phase that will make or break its future, and 5G is its final push.

My key takeaway from Helium’s successes and failures is that this model can and should be replicable in countries/regions that are just beginning to develop 5G infrastructures and should only scale when demand is proven. Mesh networking (with public access) shows significant benefits over legacy processes and should be the default way to build as we progress. Although I doubt it will have significant benefits if strong aspects of DeWi are to be stripped away, particularly on the cryptoeconomic side (aligning user actions) and permissionless nature. Telco companies can either ignore, embrace, or leverage DeWi into their workflow. By embracing DeWi, they could utilize networks like Helium or Pollen mobile to power their networks, taking them in as a neutral host. They could even start their version (that is more regulatory-friendly) where the target mining audience is a group of institutions (i.e., enterprise deployers). mimiLFG