Is Your Helium Radio Signal Good Enough?

How do you make sure a radio sig­nal is with­in accept­able lim­its for Heli­um? Short ver­sion: Assert your anten­na gain (includ­ing your cable loss) and loca­tion accu­rate­ly and you don’t need to do any­thing else.

Wait, you want more? Dawg, why did­n’t you say so? Let’s dive in!

First, let’s start with why we need an “accept­able” strength. Three words: Proof Of Cov­er­age. We need to be able to prove that we’re actu­al­ly pro­vid­ing cov­er­age where we claim we are. This is impor­tant for two rea­sons. First, when busi­ness­es hear about the world of IoT and Heli­um, if we can show them a map of where our cov­er­age has actu­al­ly been proven to exist and at what strength, they can quick­ly make a deci­sion regard­ing whether or not they want to use the Heli­um Net­work, and if they need to add a new Hotspot to pro­vide coverage.

Sec­ond, we need to prove cov­er­age in an accu­rate way in order to com­bat gam­ing, aka cheat­ing. Under­stand­ing how this works requires a lit­tle bit of radio the­o­ry, but relax, I’ll walk ya through it.

As you read this and the exam­ple I give at the end, you’ll come to the under­stand­ing that as of now, March 2022, the RSSI lim­its are gob­s­mack­ing­ly lax. This will change. For now, just book­mark that idea as “work in progress.” Onward!

Let’s start with that sig­nal strength, or RSSI. RSSI stands for Received Signal Strength Indi­ca­tor, and, as its name indi­cates, is a mea­sure­ment of the received sig­nal strength. In order to know what the received sig­nal strength *should* be, we need to know a few things. Those are bro­ken into 3 sections.

1. Beaconer Information

  • Bea­con­er’s trans­mit pow­er. In the US, most of our hotspots trans­mit at 27 dBm. Your region may be different.
  • Bea­con­er’s cable loss. (in the Heli­um app, this is includ­ed in the anten­na gain sec­tion) This is a func­tion of cable effi­cien­cy and length. 100′ of LMR400 at 915 MHz will lose 2.2 dB.
  • Bea­con­er’s report­ed anten­na gain. 

Know­ing the Bea­con­er’s trans­mit pow­er comes from know­ing what region of the world the Hotspot is in, and what the legal lim­it is for pow­er out­put. For exam­ple, US915 can blast at 27 dBm while EU868 is lim­it­ed to 16 dBm. 

2. Distance & Region

Over any giv­en dis­tance a radio sig­nal will lose a the­o­ret­i­cal amount of strength. The actu­al amount can change, some­times dras­ti­cal­ly, depend­ing on envi­ron­men­tal char­ac­ter­is­tics like veg­e­ta­tion or build­ing obsta­cles, and to a lessor extent from humid­i­ty, rain, snow, sleet, and pol­lu­tion. The the­o­ret­i­cal amount is called FSPL, or Free Space Path Loss. This is the loss in sig­nal strength in “nor­mal” clear air.

3. Receiving (Witness) Information

  • Wit­ness’ report­ed anten­na gain
  • Wit­ness’ report­ed cable loss (in the Heli­um app, cable loss is includ­ed in the anten­na gain section)
  • Wit­ness report­ed RSSI (at what strength did they “hear” the signal?)

How does that look? Lemme draw ya a pic­ture, and yes, I’ll make up a few numbers.

It starts off with the Hotspot’s Bea­con out­put, los­es pow­er along the cable, the anten­na shapes and focus­es the ener­gy out into space, where ener­gy is lost, the receiv­ing anten­na picks it up accord­ing to its gain, ener­gy is lost again as it goes down the cable and is final­ly received in the Wit­ness­ing Hotspot.

The ‑81 RSSI dBm that the receiv­ing Hotspot (aka the Wit­ness) reports is then com­pared against what it should have received at, giv­en how far away it’s assert­ed from the Bea­con­ing Hotspot as well as the out­put pow­er for that region.

Now, there’s a prob­lem, because there’s the the­o­ry, then there’s the real world, and then there’s our inter­pre­ta­tion of what should “count” in the real world. Remem­ber that term “gob­s­mack­ing­ly lax” I used above? Here’s where you start to under­stand it.

You see, as much as we’d like to think that we humans can accu­rate­ly assess and cal­cu­late the world around us, we’re not always as accu­rate as we’d like to be. The actu­al path loss over 20 km may be much high­er if there’s veg­e­ta­tion in the way (lookin’ at you, Flori­da) or build­ings (Hi New York!). Since the radio mod­els Heli­um uses don’t yet take veg­e­ta­tion or build­ings or oth­er obsta­cles into account, the actu­al results can be much weak­er than pre­dic­tions, some­times by large margins.

How does it work? Let’s take an exam­ple. Here’s Ama­teur Jade Hare with a 3 dBi HNTen­na Wit­ness­ing a Bea­con from Amus­ing Eggshell Mon­goose, which has an assert­ed anten­na gain of 5.8 dBi and is 40 km away. 

Ok, so that’s cool, but where did I get the info about Amus­ing Eggshell Mon­goose? Right off Explor­er.

That could be a 5.8 RAK, or it could be a 6 dBi from McGill with .2 of cable loss. By the way, you can add or sub­tract all this “dBi, dB, and dBm” stuff inter­change­ably with­out wor­ry­ing about the dif­fer­ences for now. Radio geeks will bris­tle at that state­ment, by the way. I guar­an­tee you I’ll get at least one snide com­ment about how you can’t pos­si­bly do that and live. What­evs.

What about the FSPL? I head­ed over to Every­thin­gR­F’s cal­cu­la­tor for that, here’s what I got:

I’ll call that 124 to make the math easy. So now we’ve got every­thing we need. 

  • Bea­con­ing Hotspot Out­put (US915) = 27 dBm
  • Cable Loss = .-2 dB
  • Bea­con­ing Hotspot Anten­na Gain/Cable Loss = 6 dBi
  • FSPL = ~124 dB
  • Receiv­ing Anten­na Gain (includ­ing cable loss, in this case almost noth­ing because it’s 4′ of LMR400) = 3 dBi

So what SHOULD the report­ed RSSI be?

27dBm — .2 + 6 — 124 + 3 = ‑88.2 dBm RSSI

What was the report­ed RSSI? ‑108 dBm! 

That’s almost a 20 dB dif­fer­ence, and it still cleared the line! What does that tell you? It tells ME that it is bloody dif­fi­cult to cor­rect­ly assess loca­tion based sole­ly on RSSI and cur­rent radio mod­el­ing. I know for a fact that AJH is where it says it is; it’s my Hotspot. I don’t know about Amus­ing Eggshell Mon­goose, but if it’s a gam­ing hotspot it’s doing a ter­ri­ble job of earn­ing. My guess is that AEM is where it says it is.

Where does this leave us? With a thornier prob­lem then when we start­ed, as we now real­ize that gam­ing through atten­u­a­tion is sig­nif­i­cant­ly eas­i­er than you might have thought, since the lee­way is so great. For now, at least you know how to check your RSSI val­ues. If you want to see the equa­tions Heli­um is using, check their github, start­ing here. Yeah, it ain’t easy to read. 

The solu­tion is for Heli­um to start ramp­ing up the fspl_factor chain vari­able, which will clamp down on allow­able RSSI and make it more of a real con­trol. Before you start scream­ing about what Heli­um should do and how fast they should do it, keep in mind that they are man­ag­ing an enor­mous and com­plex sys­tem, and gam­ing is (these are MY words, not theirs) a rel­a­tive­ly small prob­lem com­pared to keep­ing the blockchain running. 

The great news is that if you’re not an RF geek, none of this mat­ters. Assert your cor­rect anten­na gain includ­ing cable loss, and focus on what’s actu­al­ly impor­tant, which is WHERE you put your hotspot. Need help with that? Take my Heli­um Basic Course or the Heli­umVi­sion Mas­ter Class; I built ’em to help you under­stand the things that actu­al­ly mat­ter in Helium.

Rock on!

p.s. Giant thanks to Jere­my Coop­er for his help explain­ing this to me and fact check­ing my usu­al hasty assump­tions. All mis­takes are mine, all right­eous accu­ra­cy is his. If you want to get an idea of the expe­ri­ence and skill being thrown at this prob­lem, I strong­ly rec­om­mend you check out the inter­view we did on YouTube. 

Additional Ultra Geeky Thoughts from @Jerm

Although most US hotspots can and do use a con­duct­ed pow­er of 27 dBm when trans­mit­ting, a few things can make it different:

  • Man­u­fac­tur­er uses a radio that can’t out­put 27 dBm.
  • The blockchain has asked the radio to trans­mit at a low­er pow­er due to EIRP lim­its, but the radio can’t do that spe­cif­ic pow­er, so it choos­es the next low­er lev­el. Exam­ple: A 10 dBi anten­na in the US would require a +26 dBm trans­mit. If the card can’t do +26 dBm it might do +25 dBm.

When either of these occur, the actu­al trans­mit­ted pow­er the hotspot used is report­ed on the blockchain. This means that the aver­age user just rely­ing on Explor­er may not see it, but the pre­cise data is entered (and used) by the blockchain.


24 responses to “Is Your Helium Radio Signal Good Enough?”

  1. Thank you for the great analy­sis! I’m in the mar­ket for a new anten­na, so this was incred­i­bly helpful.

  2. Hi Nathan
    I am going to set­up around 10 Sense­Cap in my town, have pos­si­bilty istall them around 20 meter high , all of them ( that the plan) with clear view, and all see each oth­er. As they not yet arrived , i am study­ing as much us i can…Here is my ques­tion. Plan is to set them up , each one 1km each oth­er with 3dbi anthena.
    Doing a math and RSSI is around 67
    Sense­Cap 14 dB — 1db (cable loss) + 3 (anten­na g) — 85 dB ( fspl ) + 3 (anten­na from oth­er side) — 1 (cable loss) = 67 dBm RSSI 

    Look­ing at oth­er simil­lar set­up RSSI most of them around 120 —

    Am i doing some­thing wrong , is that RSSI will cause witness_rssi_to_high ?
    Greet­ings from Poland, amaz­ing webpage!

  3. Hi Pawel, I think you’ll be fine, keep me post­ed on how it goes! ~Nik

  4. By the way, you can add or sub­tract all this “dBi, dB, and dBm” stuff inter­change­ably with­out wor­ry­ing about the dif­fer­ences for now.”
    Sub­tract­ing pears from apples the cal­cu­la­tions will always be wrong.

  5. I don’t believe so in this case, but help me understand.

    For exam­ple, if I have a min­er putting out 27 dBm hooked to cable that has a loss of 2 dB con­nect­ed to an anten­na that has a gain of 5 dBi, then the “anten­na gain” I enter into the Heli­um app to let the Net­work know what to expect will be 27 — 2 + 5 = 30 dBm.

    What am I missing?

  6. Nik, my under­stand­ing is that in your case, you enter 3db (5–2) in the Heli­um app set­up, along with your anten­na height. They know your pow­er output.

  7. Help me under­stand how Ama­teur Jade Hare can bea­con and wit­ness like this at 3dbi? I am guess­ing ampli­fi­er and atten­u­a­tion by not report­ing actu­al anten­na dbi?

  8. It’s a func­tion of the loca­tion. It has a tremen­dous­ly good line of sight. There’s not more than that to it; no fan­cy ampli­fi­er or atten­u­a­tion etc. It’s just a 3 dBi HNTen­na placed at about 6′ high on top of a 3,000 mtn.

  9. Thank you for respond­ing. I appre­ci­ate what you do.

  10. Hi there. Very good con­tent. In the cal­cu­la­tions of the arti­cle I think that you are tak­ing into account the anten­na gains twice. The FSPL cal­cu­la­tor is already using them, so no need to sum them again. Or you could just set them to 0 in the cal­cu­la­tor. The the­o­ret­i­cal RSSI of your exam­ple should be ‑88. Tell me if I am wrong. Regards,

  11. Nice catch, that makes sense. Let me dou­ble check with my radio geek homies to make sure, then I’ll update this. Thanks Jonny!

  12. Hi, very inter­est­ing. I’m won­der­ing if you can help me under­stand an invalid wit­ness report I got a cou­ple days ago.

    I’m up lit­tle over 100 meters, and most of the oth­er sites I can see are about 20km away on the oth­er side of San Fran­cis­co Bay. I put up an 8 dBi anten­na a few days ago, with 15 feet of LMR-400 on a Bob­cat 300, rat­ed loss through the cable of 0.7dB, so I’m report­ing 7.3dBi. For the most part it’s work­ing well, but there are a cou­ple of invalid wit­ness reports that don’t make sense to me. Exam­ple: the oth­er sta­tion is a Cal-Chip Con­nect­ed Devices some­thing-or-oth­er, report­ing its loca­tion as 33km away, report­ed as 30m up (prob­a­bly a mis­take for “above ground”, but close to cor­rect for ASL height), Explor­er shows ~33km dis­tance (mea­sured 33.39km), with a 5.8dBi anten­na. I get an FSPL of 108.94434015dB from the cal­cu­la­tor. Assum­ing 27dBm and zero cable loss for the trans­mit­ter, their EIRP would be 32.8dBm, and 32.8 — 108.94434015 is ‑76.14434015. Explor­er tells me the trans­ac­tion had an RSSI of ‑80dBm, which would make sense if we add some cable loss on the trans­mit side, and an SNR of 7.8dB. The SNR does seem high, but–maybe we just got lucky, and the noise hap­pened to be low?

    edit [Just real­ized I neglect­ed to add my anten­na gain into the final RSSI cal­cu­la­tion, which brings it up to ‑68.84]

    Any­way, I don’t under­stand why this was flagged as invalid. Can you explain? Also, I’m very con­fused as to the cur­rent state of PoCv10, 11, etc., and frus­trat­ed at how dif­fi­cult it seems to be to get defin­i­tive, offi­cial infor­ma­tion. Com­ments on that?

    Thank you!

  13. Apolo­gies for being so chat­ty. As Jon­ny not­ed above, the “FSPL” cal­cu­la­tor is not real­ly cal­cu­lat­ing FSPL, it’s doing path loss, includ­ing anten­na gains. True the­o­ret­i­cal FSPL then would be 122.0521234dB. With that cor­rec­tion, my pre­dict­ed RSSI comes out to ‑81.95212337dBm, so very close to the actu­al val­ue. If the trans­mit­ter’s anten­na gain is actu­al­ly 8dBi, and my cable loss is just a bit high­er, it’s almost an exact match. In any case, the num­bers are rea­son­able, and both sta­tions are with­in legal lim­its for the US.

  14. How much are your services?

  15. Hi Toby, you can find all ser­vices & prices via the tabs on the home page.

  16. Vilmos Opra Avatar
    Vilmos Opra

    Hi Nik!

    I just got a kid­ney trans­plant, your inter­est­ing writ­ings will heal me! When I go home I will set up the min­ers again based on your descrip­tions! Greet­ings from Europe, a small coun­try, this is Hun­gary! Vil­mos Opra / no Oprah/

  17. Great infor­ma­tive site — thanks for shar­ing all this knowl­edge! A quick ques­tion if I may. If I have a 6dBi anten­na with a 2dB loss cable is that the same has hav­ing a 4dB anten­na with zero loss cable (hypo­thet­i­cal­ly obv!)

  18. It is as far as sig­nal strength. It’s not as far as sig­nal pat­tern. Does that make sense?

  19. Thanks for your reply. Yes it does make sense, that was what I was get­ting at. I was won­der­ing if mount­ed on my roof and­my 6.2 dBi anten­na with a 30ft LMR400 cable would behave like a 5dBi anten­na in all respects! So, my anten­na is rat­ed as 6.2dBi and I entered 6.2dBi in the app when I assert­ed my hotspot as I assumed that was the cor­rect thing to do. Assum­ing a cable loss of 1.2 dBm should I edit my hotspot to show a 5dBi anten­na as that is rough­ly what it is I guess…?

  20. Yep, exact­ly right. You’re aim­ing to let the net­work know what to expect as far as sig­nal strength.

  21. Cool, thanks again for tak­ing the trou­ble to answer. I am in the UK and I only set­up my first hotspot 10 days ago, I bought a 6.2 dBi anten­na a week ago and put it on my roof. My stats are 39 wit­nessed in three days and 103 bea­cons 7 day aver­age accord­ing to Heli­um Explor­er. I am won­der­ing though if this is like­ly to be improved by a low­er gain anten­na as I have fair­ly good lines of sight for about 15/20km in a 270 degree radius and line of sight for about 7km for the oth­er 90 degrees. The biggest con­cen­tra­tion of oth­er hotspots though are in the 270 degree arc about 10km away and a lot of peo­ple seem to just plug them in and put them on a win­dowsill by the looks of the stats of some ran­dom ones I have clicked on, so would a low­er gain anten­na be more like­ly to reach these? When I say low­er gain I mean one of the 3dBi ones. All the dia­grams and expla­na­tions seem to sug­gest that a low­er gain anten­na is bet­ter in a block of flats, which is fine, but oth­er hotspots are like­ly to be only meters away in that sce­nario. How does a low­er gain anten­na fair when the lines of sight are good to the next town but it is 15km away..?

  22. Hi James, the only accu­rate results will be found from test­ing, but at a guess I would­n’t expect a huge improve­ment from chang­ing your anten­na; they just don’t make that much of a difference.

  23. Hi to everyone.

    I hope you guys have a won­der­ful day and hope­ful­ly some­one can help me a bit.

    I have 3 anten­nas. 8 dbi, 6 dbi and 4 dbi. High­ts of the anten­na is around 20 m high (Lodnon/UK) I have a 5 m cable LMR400 and also 10M cable LMR400. Min­er is bob­cat 300. What I am won­der­ing is this. With 8 dbi, I get so many invalid wit­ness because they’re too close. With the 6 dbi not so many and with 4 almos none. So my ques­tion is why does this 8 dbi have so many invalid wit­ness? Should it not reach those min­ers that are a bit out­side of lon­don? I thought if I use the 4 dbi, it will not leave the cen­tral lon­don where I live so I got the 8 dbi to get out of lon­don and wit­ness those that are over 10 km away. My oth­er ques­tion is what if I con­nect two anten­nas to the same min­er. Will that work? Has any­one tried it? The 4 dbi and the 8 dbi. The 6 dbi I have does fine on its own.


  24. Hi Tony, there are a few answers that are com­plex regard­ing how RF works, but one use­ful very sim­ple one: Choose either the 4 or the 6 dBi, sell the rest, and work on oth­er more pro­duc­tive aspects of Heli­um. Opti­miz­ing anten­nas is a fun game for geeks, it’s not real­ly impor­tant for the rest of us. :). Let me know if that helps.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.