Posted

Pollen Mobile, a product of the Pronto.ai self driving car startup engineering team, burst onto the decentralized wireless scene with a small San Francisco Bay Area launch in February 2022. In Part 1, I established that the Pollen Mobile impressed me by delivering the core product after a super fast development cycle started in late 2021. Given the product is based in CBRS technology and has a working CBRS-based Proof of Coverage mechanism, the Helium community was quick to compare – and judge. In this Part 2, I’ll break down the differences, address the implications and tackle my clickbait headline: is Pollen a Helium 5g killer? (According to Betteridge’s law of headlines: no)

How are they the same?

Helium 5G, despite the name, is firmly rooted in currently available CBRS radio technology underpinned by the 4G LTE protocol (this will evolve as available tech evolves). Pollen shares the exact same mobile hardware & software stack: CBRS radios and Magma packet core. So, both products are able to service modern cell phones: iPhone 11+, Pixel 4+, Samsung s20+. Operational experience in planning deployments, tracking signal propagation and measuring coverage is more or less identical between the products. In this sense, Pollen is an excellent preview of what Helium 5G stands to offer. Both products have set a price of $0.50 per gigabyte of data usage, a rate between 70-90% cheaper than what incumbent MNOs may charge MVNO partners for roaming. Finally, and very important, both projects offer or plan to offer a Proof of Coverage (PoC) mechanism for their CBRS/5G coverage product. Referring back to Part 1, PoC is the rocket fuel for network buildout.

The token economic model also carries some similarities. Helium allocated about 35% of tokens for insiders, Pollen has a similar quantity, and both are subject to lockups & cliffs. There are funds dedicated to community development grants and there is a mechanism to charge for data consumption. However, I argue the differences outweigh the similarities.

Where are they different?

The most glaring differences between Pollen Mobile and Helium 5G come down to foundational societal value offering. Pollen is actually much closer to Helium LoRa than it is to Helium 5G. Helium LoRa started by building a fragmented and disparate LoRa network and achieved roaming status after demonstrating massive scale through real coverage. Helium 5G is taking the opposite approach as Helium LoRa: roaming-first, also known as neutral host.

An obvious criticism for the Pollen roadmap is the lack of market participants, since they have no neutral host plans (yet). Who wants to bother trying to access a wireless network which is admittedly tiny? I’d like to remind readers Helium LoRa was in an identical predicament just two short years ago. (Side note: Pollen has an eSIM mechanism which rewards you every day you connect to the Pollen network with about 75PCN as of today. There is no such mechanism being planned in Helium 5G outside of witnessing.)

With Helium 5G’s neutral host plan, any random individual can roam onto a Helium 5G hotspot with their ordinary cell phone plan, so the total addressable market is enormous from day 1. However, neutral host is like the final boss of network buildout. Striking an agreement with AT&T, T-Mobile and Verizon before having a large mature network is a monumental task. What helps to build large networks from scratch? Proof of Coverage! One problem. The Helium 5G proof of coverage roadmap is significantly delayed, deprioritized in favor of neutral host work, but also subject to changes in governance & blockchain roadmap. Without a pivot, there is a chance we won’t see Helium 5G PoC this year. By skipping neutral host and placing CBRS PoC up front, Pollen borrows Helium LoRa’s scaling strategy and does away with Helium 5G’s largest development hurdles. There are people who think Helium 5G doesn’t need PoC to succeed, and they could be right. I don’t want to call the delay in Helium 5G PoC a blunder just yet, but the growth of Pollen may change my opinion.

Currently the Helium 5g Proof of Coverage roadmap claims a testnet by June 2022. Add a few months of margin for validation and that puts the PoC release date somewhere in Q3 or Q4 of 2022. It’s reasonable to say Pollen PoC has beaten Helium to market by between 6-8 months.

Besides the approach to market, Pollen & Helium differ significantly in blockchain selection. Pollen adopted the Solana SPL token whereas Helium is of the era without viable L1’s (pre 2020) so had to create its own L1. This foundational difference has cascading effects, especially when it comes to interplay within emerging DeFi infrastructure.

Finishing with economic differences, Helium is clearly more mature in its economics than Pollen. Helium has several drivers of hodling: validators and (soon) token-lock based governance. Pollen governance, validation, etc, is nebulously dealt with in Phase II and III. Not to mention PCN isn’t even available to trade anywhere. Finally, and most significantly, is how Pollen and Helium address the question of utility. Helium has a fixed-price data credit (DC) system for usage and a floating price token HNT for rewards/validating/voting, whereas Pollen’s one and only token PCN has a notional fixed-price from the start. I believe this is a design decision meant to avoid early conflicts with securities law. After sufficient decentralization, a new system could be voted on which would allow PCN to float. That said, right now there is no detail provided for how a fixed-price PCN token is going to function because that buy/sell arbitrage bot doesn’t just happen on its own.

Decentralization

There is a stark contrast in levels of decentralization within these two projects. Helium continues to impress me with their efforts to remove central coordination from Helium Inc and instead offload responsibilities with the Decentralized Wireless council, or with the manufacturers themselves. Recent examples of this being the requirement for manufacturers to produce and maintain their own hotspot apps and move away from a single Helium app. This is a great move to thrust more responsibility with hotspot makers and also create an opportunity for them to differentiate their product.

Pollen, by virtue of their speed, appears entirely centralized in nature. Much of the inner workings of the project are opaque: lack of API, calculations, smart contract, PoC data. However, in their whitepaper they lay out a path towards transitioning ownership into a DAO type entity. I am confident each concern will be addressed in due time.

From where I stand, Helium is striving for pure & credible decentralization with an ethos resembling Ethereum, whereas Pollen is the scrappy newcomer who has prioritized speed above all else. There are many such dichotomies in the L1 space where teams take significant liberties with the label of “decentralized”, or make conscious decisions to trade decentralization for speed. Perhaps such a scenario is evolving within the decentralized wireless space. There is absolutely room for both approaches and the L1 landscape is proof.

They actually need each other

Its easy to view these projects as competing head to head, but in such a nascent industry I firmly believe the success of one bolsters the success of the other. A project without competition is probably not a great idea to begin with, so competition first and foremost is excellent conceptual validation – it reminds me of the dancing guy analogy. I can see a world where a successful Pollen private LTE network may actually drive customers away from MNOs in dense urban areas, so in response, the MNOs are forced to compete on price and become more likely to explore creative neutral host ideas. Or, perhaps a different outcome: Helium 5G develops private network eSIM functionality and effectively swallows the Pollen addressable market (this function is on the roadmap but lower in priority). Either way it ends up, competition legitimizes an industry and I’ll be eagerly watching & participating.

Author

Posted

Pollen Mobile quietly launched on February 1st 2022 to little fanfare compared with typical crypto blitz marketing campaigns. Unless you’re a close watcher of the CBRS / decentralized wireless space, you probably missed it all together.

Part 1 of this article covers the basics of Pollen. Part 2 will address the question everybody has been asking: is this a Helium 5g killer? Be sure to follow me on twitter for more content.

Pollen immediately caught my eye so I dug deeper with an eye towards red flags: sleazy kickstarter vibes, huge markups or just ridiculous vaporware. What I saw was a legit Baicells CBRS hardware lineup sold at-cost, a drop date 7 days after announcement, claims of a working mainnet via Solana SPL token, a working Magma packet core, working eSIM and a working (!!) Proof of Coverage mechanism. No kickstarter vibes, no overpriced hardware.

Learning about CEO Anthony Levandowski’s involvement was icing on the cake. Anthony has been accused of many things but ‘visionary engineering leader’ is one label for which he won’t need a pardon. Pollen, you had my curiosity but now you have my attention.

Securing equipment

Given that Pollen’s first launch geography is my backyard, I just had to bite. While the price I paid for an indoor and outdoor unit is steep for any hobbyist budget I justified it with the thought that experience on this system translates directly to the Helium 5G rigs.

After playing with the project this week I’m here to say, YES, Pollen delivers on all of the promises. I am told this has been a big focus of the scrappy Pronto.ai engineering team since late 2021 and I am extremely impressed at what the team has accomplished in a short period of time. Larger teams have accomplished far less with far more time.

Better to tell my neighbors this is a cellphone tower or a cryptocurrency miner?

The Pollen idea

Pollen brands itself as a private mobile network: cheap, secure, private 4G/5G data for the masses. The value offered is an intersection between neutral host, home wifi and private networks. Picture a way to supplement poor mobile coverage at your home, or an eSIM you can install on your phone to feel confident you are getting rock bottom data rates when you’re out in public.

However, the way of achieving this dream sounds a little strange if you’re new to decentralized wireless. Building a network, one small owner-operated roof-mounted radio at a time? How can you compete with ATT, T-Mobile and Verizon with their massive towers and spectrum? This is where it helps to know about Helium and the effect Helium’s Proof of Coverage (PoC) incentive has had on the human psyche. Helium has created a culture of people willing to buy hardware, climb roofs and hustle for locations all to the benefit of the Helium network. In other words, Helium PoC convinced participants to absorb the largest costs of building a mobile network in exchange for an equitable portion of the network value. Needless to say, proof of coverage is an effect to behold. One could dismiss PoC as empty crypto ponzi-nomics but you can’t argue with the results.

Here’s what I want to emphasize: Pollen is Helium – but replace the LoRa network with a 4G/5G network. Climbing the roof is the same, hustling for locations is the same and proving coverage is the same. As I’ll explain in Part 2, Pollen adopted Helium LoRa’s strongest growth strategies and skipped Helium 5G’s largest development hurdles, however their private network goals may put a ceiling on success.

Does Pollen work?

Enough talk, does Pollen work as promised? YES! My founders kit consisted of a “Flower”, or the Baicells radio and Magma gateway pair, several small embedded Beaglebone linux devices with a custom printed circuit board and Band 48 modem (Bumblebees) and a bunch of eSIM downloads (Hummingbirds).

The Pollen token PCN can be earned in several ways. eSIM devices can earn a small amount per day by simply attaching to a Pollen CBRS Flower node. Bumblebees earn in a similar fashion, except earnings are multiplied according to the number of unique geospatial Hex9 attachments in a day, coordinated by GPS (side note, this is an abuse vector). Most earnings will go to the Flower, who will receive 10x more rewards for each unique Flower-to-Bumblebee Hex9 confirmation, up to a maximum dictated by the type of Flower.

Just two days after securing equipment, I’ve seen all the main functions work as intended. I’ve proved my location with bumblebees, I’ve roamed with eSIM devices, I’ve had a CPI come check out my install, and earned PCN. I am thoroughly impressed.

The filled turquoise cells indicate a bumblebee has verified a radio connection within this hex. The empty cells indicate…something else (i’m not sure).. There is also a bounty system, where signal validation events in some cells are worth a bonus. Clearly there is much to be improved in the UX of just about every feature, but the core is here.

At a later date mobile data transfer & other mechanisms will factor into rewards. Worth emphasizing is that for now the network is basically dishing out stable coins because the PCN token is designed to be pegged at 1PCN = $0.10. To me this looks like a trick to avoid securities law conflict, akin to airline miles. Anthony, take note: compliance will be a significant part of your life if Helium’s Amir Haleem SEC experience is any indication. People love speculating, DeFi and price action. Give the people what they want.

Can Pollen work?

Can Pollen achieve its goal of creating supplementary offload opportunities for eSIM users? There are ~40 Helium hotspots per square mile in San Francisco, which is a ridiculous density. If just 10% of those users deployed a CBRS radio, that’s 4 radios per square mile. Suddenly a viable cheap & private mobile network in San Francisco doesn’t seem crazy. The early adopter culture of crypto is legendary, so I actually have no doubt we’ll see some serious adoption in SF at least. However, I think Pollen token economics need a major overhaul before we can expect stratospheric results.

Above is a modeled CBRS Band 48 coverage map of an outdoor node similar to mine, positioned in prime real estate overlooking Dolores Park, San Francisco. On a sunny day, this park sees tens of thousands of folks visiting and hanging out for hours. One of these outdoor nodes can easily blanket the entire park with coverage. The thought of giving visitors an alternative source of mobile data is very enticing; I fully expect we will see these kinds of deployments pop up, owned by the lucky few with access to such valuable real estate.

I found these signal propagation models quite accurate when applied to my own neighborhood. After doing spot checks with my Pixel 6 Pro + eSIM I was able to stream youtube videos about .25 miles from the base station, within a total area of about a tenth of a square mile filled with about 80 suburban homes. I lacked line of sight for pretty much the whole time with tons of large trees and typical residential objects blocking LOS. Superior positioning will surely grow this usable area by a factor of 2-5x.

In summary:

  • The good: Super early no-nonsense release, all the major functions work, the tech & packaging is downright cool.
  • The bad: everything is clunky, nonexistent tooling or ability to understand status/function.
  • The ugly: broken token economics and a long road towards credible neutrality & decentralization.

No matter how you slice it, a talented team just entered the decentralized wireless ecosystem who have managed to pull off something exceptional in a short amount of time. 2022 is starting off to be a great year for DeWi. Stay tuned for Part 2 where I’ll go into depth about how Pollen differs from Helium 5G.

Author

Posted

HIP20 is the present-day monetary policy for Helium and utilizes tried and true crypto economic mechanisms: capped supply and halvings (github). Thrown into the mix are the concepts of Burn & Mint equilibrium and Net Emissions. Understanding the network outlook boils down to answering one simple question: are the total number of HNT growing or shrinking? Despite the clear language of HIP20 it’s still not easy to form an intuition for this question. “$1M/month of network usage means a price of $20” is a line popularized by JMF, but what does it mean? How does one arrive at this conclusion? Are there other surprising conclusions?

→ follow me on twitter @DeWiGoSite

This post is really just an explanation of key concepts I wished were around when I tried to digest HIP20, starting with some of the common phrases thrown about when discussing Helium economics.

DC Burn is the core function of the network, and what we’re all here trying to do. Less understood is that there are two types of DC burn functions and the accounting is different for each.

  • Data transfer (not deflationary, up to a point)
  • Non-data transfer (deflationary)

Since the value of a data transfer DC burn is paid to the hotspot routing traffic, by definition it does not deflate the total supply of HNT until the total DC transfer in a 30 minute period (epoch) exceeds the allocation codified in HIP20. Today, this amount sits at 35% of total HNT emission per epoch, or approximately 1712 * 35% ~ 600 HNT. Due to nascent network utilization, true consumption is less than a 1/10th of a single HNT per epoch, so the unused portion is generously given to PoC participants.

Non-data transfer DC burns are from actions such as HNT transaction fees, onboarding fees and location assert fees. Since nobody is the recipient of these fees, there is no connection with Burn & Mint, and they are by definition purely deflationary.

Burn and Mint is where the use of the network is derived from burning the native token to redeem what amounts to non-transferable airline miles. The Multicoin Capital post will do the explanation more justice than I ever can. To say B&M is “kicking in” is really just a situation where demand-side HNT burn (constant cost) exceeds supply-side hotspot reward allocation (fluctuating), and thus the total supply of HNT is under deflationary pressure. Equilibrium is once again reached when the native token appreciates in price.

Put another way, when B&M is “kicking in”, hotspots won’t actually get paid the nominal rate for data transfer ($0.50 per 5G gb, $0.00001 per LoRa packet), instead earning a scaled portion. It always costs the same to use the network, but data routers only earn up to parity with cost. People assume this data transfer rate is fixed forever, but high network demand paired with volatile sentiment on $HNT can result in data transfer earnings less than expected. Today, it’s a far-off scenario but if the network really catches on it will surely happen and possibly catch people off guard.

Halvings. Like bitcoin, the monetary equation for Helium changes on a regular basis. Helium emission rate will halve every two years, but actually goes through changes on a yearly basis when the ratio of reward allocation between PoC, data transfer and founder reward all change. However, until DC burn from data usage becomes a greater quantity of usage, yearly changes are not especially relevant. I expect further changes to this reward schedule as more HIPs get implemented which factor in new forms of Proof of Coverage.

Net Emission is relevant many years from now and is framed as ‘a mechanism to ensure network participants always get paid’. The value of Net Emissions is locked at 34.42 HNT per epoch, and will not go down during halving cycles. Another way to think about Net Emission is it establishes a non-deflationary network utilization floor, or, a forever-window of 1:1 payment for data transfer participants.

There is a future date at which Net Emission, 34.42 HNT per epoch, makes up the lion’s share of all HNT available. Today, Net Emission consists of only 2% of the bucket per epoch (34.42 / 1712 HNT), in 5 years it will be 8%, but in 10 years it will be much higher at 65% and eventually consist of close to 100% of all available awards.

Crypto moves fast – how many headline projects from 2017 are around today? Its an interesting economic mechanism and critical for the long term network utility, but not likely a factor in Helium’s near term survival.

Painting a picture

I present below the first true visualization of HIP20. Using two simple numbers – price of $HNT and daily DC burn, you can use the chart to determine if the maximum supply of Helium is increasing, decreasing, or staying the same. Because the economic model changes on a yearly basis, so must this graph, so the present-day 2021 3 Year graph is next to the 20 year steady-state chart.

Make note of the location of lines representing DC B&M Equilibrium and N.E. Equilibrium. They actually switch places in the year 20 chart because halvings decimate HNT total emission, allowing Net Emissions to become the steady state non-deflationary floor. Year 3 also contains a section I called Slowed Inflation, which amounts to contributions to total supply from PoC. By year 20 this inflationary buffer has disappeared entirely and the inflationary floor shrinks to meet the line of net emission.

This brings us back to JMF’s million-dollar comment: “$1M DC burn per month (33k per day), guarantees support at $20”. You can see this statement visually now by tracing the $20HNT dashed line and seeing it intersect with the $33k per day DC burn price. This relationship is linear, meaning $2M/mo burn equates to $40 equilibrium, $10M/mo burn to $200 equilibrium and so on.

A less sexy but more accurate statement would be “In Helium’s steady state economic environment 17 years from now, DC burn in excess of $1M per month causes deflationary pressure on total supply when the token is priced below $20”.

(But that’s not as good of a tweet.)

I must point out this pertains to Data DC Burn only. The network burns on the order of $200k worth of onboarding fees daily, which is indeed deflationary. You can use the non-data DC burns alongside these charts, but the colors and boundaries will lose meaning.

The Helium economics are a simple but cleverly designed, utilizing well understood and accepted ideas within the crypto world along with a few new ones. The long term viability of the project requires the existence of a healthy stream of demand-side HNT burn. We are NGMI as long as HNT demand is driven by circular onboard and transaction fees. Stay tuned for a future article where I will discuss those exact prospects as it pertains to 5G.

Author

Posted

Helium has 240k+ hotspots globally, growing at a rate of 50-70k monthly. It’s already the largest contiguous wireless network in the world, with several roaming deals under its belt and more in the pipeline.

The question remains: how much coverage does Helium really provide?

Spoiler alert: if your town has a Starbucks, theres a good chance you already have coverage.

→ follow me on twitter @DeWiGoSite

For the United States, I can answer this in a quick and dirty way using geolocation, zip codes and census data. This can be done easily with uszipcode and geopy python packages. First, I resolved the zip code of all online US hotspots based on their latitude / longitude (about 89k as of writing), then queried basic zip code statistics like area and population. It’s obviously not a perfect approach, but zip-basis gives a more meaningful picture on coverage than the city-basis already out there. Unpopulated or “weird” zip codes are excluded (such as military bases, large uninhabited land, territories, special government designations).


I should mention there are DeWi grants which will do similar things in a more robust way by using actual PoC receipts and more sophisticated mapping. I’m excited to see what they come up with.

Definitions

Each zip code will be classified into the following buckets based on population divided by area. Using DoD definitions:

  • Urban zip codes have > 3000 people per square mile (97 million people live in this criteria)
  • Suburban zip codes have 1000-3000 people per square mile (71 million people)
  • Rural zip codes have < 1000 people per square mile (143 million people)

Its worth reiterating that I’m only counting online hotspots. The online hotspot rate fluctuates between 75-85% based on a variety of factors.

Zip code coverage

Above is a graph of total US zip codes containing at least 1 hotspot. Frankly, it’s an astonishing graph — Helium is spreading across the United states like a living organism. Below shows each demographic broken out over time. Last year, Helium had a presence in 40% of urban zip codes, today it’s 90%+. According to the rules of HIP17, when hotspot density grows too high the earnings go down. We are seeing this play out in real time as suburban zip codes are now the fastest growing demographic, with rural areas showing notable growth after the middle of 2021. By 2022 the rural zip codes will be the only remaining areas for large profit potential.


Class Total US zip codes Zips with hotspots %
Urban 3493 3149 90%
Suburban 3436 2500 73%
Rural 25958 4091 15%

(Note: Table dated October 2021, will be out of date very soon)

Node density

It’s also helpful to look at coverage according to density, defined as the number of hotspots divided by area of zip code. This is a good raw measure of people living in covered areas.

Hotspot density Number of zip codes Population of zip codes
> 0 hotspot/sq. mile 9740 229 million
> .5 hotspot/sq. mile (1:hex7) 4368 118 million
> 1 hotspot/sq. mile (2:hex7) 3178 87 million
> 2 hotspot/sq. mile (4:hex7) 2074 57 million
> 4 hotspot/sq. mile (1:hex8) 1158 33 million

(Note: Table dated October 2021, will be out of date very soon)


For the Helium crowd obsessed with hex placement, according to the area mapping table, a single hex7 is about 2 square miles, and a single hex8 is about .25 square miles. There are 118 million people live in areas with one hotspot per hex7, which is decent coverage with some redundancy. One hotspot per hex8 is terrific coverage and 8x more dense than 1:hex7, with about 10% of the United States living in these conditions. In reality, “true coverage” is better than these figures since a single well placed antenna can cover dozens of square miles.

The Helium Explorer shows hex8 cell size. One hotspot per hex8 would make the explorer cell map completely green, just like San Francisco, Los Angeles, Chicago or NYC.

Arguably, Helium is about 30% of the way to the end goal of 1 hotspot per square mile in the United States. By the time you read this, the data will already be out of date as the network marches towards 10x growth by 2022! Simply amazing.

Revised October 20th 2021 to fix some calculation and formatting errors.

Author

Posted

A clickbait headline admittedly, yet the explosive growth of the Helium network is not unlike the spread of a powerful meme or a contagious condition.


→ follow me on twitter @DeWiGoSite


Cryptocurrency has always been a solitary endeavor: stake in a pool, operate GPUs, trade the market, or click-to-compound in a yield farm (reminiscent of swarmsim.)


Helium is different. Helium brings to crypto the #1 rule of the oldest asset class: location location location. As I recently pointed out, well placed nodes are 40x more productive than poorly placed nodes. This effect creates huge pressure for an enterprising individual to work his or her network to spread out geographically.

Viral MLM

Helium operators have all experienced it. You just finished an install, then as if by instinct, the host states: I can get more locations. What’s my cut? Congratulations: you are now a Helium superspreader. And believe me, the idea of Helium is contagious. It rings in your head like a bell. It’s like learning about bitcoin for the first time. The simple rules dictating Proof of Coverage produce amazing emergent social behavior — it’s an effect to behold.


These days, everybody knows the measure of a disease virality is the basic reproduction number (R0), defined as the average number of people infected by a contagious host. According to the NIH, COVID-19 R0 is between 2-3 by most estimates. What is the basic reproduction number of Helium? Let’s take a look.

Getting quantitative

Calchip and Bobcat have disclosed the average order being about 2.5 to 5 nodes per order. At the time of writing, there are 48k wallets and 146k hotspots, for an average of 3 hotspots per wallet. If we take the average of these numbers as the reproduction number, then we wind up with a number on par with COVID R0 estimates. So perhaps in a very crude way, we can answer “The Same” to the question posed in the title :)


The full distribution of hotspots & wallets sheds more light. Over half of all hotspots are owned by wallets containing less than 10 hotspots, who I dub the weekend warriors. The top 4.5% of wallets are the superspreaders, with between 10 and 70 hotspots to their name. (Yes, I know it’s possible for one person to create many wallets, but I argue the current tooling and hotspot onboarding experience makes this infeasible.) The heavy hitters, ie Hosting companies, own about 25% of all hotspots.

Class # of hotspots % of wallets % of hotspots
Weekend warriors 1 to 10 95% 55%
Superspreaders 11 to 70 4.5% 20%
Hosting companies 71+ 0.5% 25%


Helium Inc has confirmed 500K preorders across 50 manufacturers, and 2 million in production. Frank Mong, COO, quoted a chipset backorder as high as 3 million in a recent interview. For each order, that’s a cascade of retail users exposed to Helium and possibly crypto for the first time. Helium is like Alexa meets viral multi level marketing, except instead of selling lotion and vitamins, we’re building something obviously useful.


The idea of Helium just clicks and is sweeping through retail like wildfire. Watch out world.

Author