search

dewi-alliance / hplans

lorawan regional map
9 stars 10 forks source link

Australia and New Zealand Switching to AS923-1 #28

Closed lthiery closed 1 year ago

lthiery commented 2 years ago

On December 10th, 2021, during the third meeting of DeWi’s LoRaWAN Working Group, members voted to initiate a change of frequency plan for Australia and New Zealand from AU915 to AS923-1. The vote count was 3 for and 2 abstentions. The video and minutes of the meeting are available.

On February 8th, HIP45 was formally approved and this PR formally initiates the change under that framework. This PR begins the minimum of 4 weeks period for written discussion. After this period, if formal dissent to the change is lodged, a virtual town hall will be held. If there is no resolution, an on-chain vote will be held where one hotspot in the affected region is given one vote.

Discussed reasons for switching to AS923-1

Discussed reasons for not switching to AS923-1

Australian-Hnt-collectors commented 2 years ago

Not Being rude here but how does this help those whom are building the network and actually obtaining government tenders which are based on AU915 and it stipulates to use that network freq for the smart farms initiative or other government procurement's which also include the AU915 stipulations with no deviations from this freq, it does not thus meaning that a lot of time, money and effort is down the drain and back to the drawing board to get the different states of Australia to accept AU923, which is a whole new mountain to climb.

jsloane commented 2 years ago

On December 10th, 2021, during the third meeting of DeWi’s LoRaWAN Working Group, members voted to initiate a change of frequency plan for Australia and New Zealand from AU915 to AS923-1. The vote count was 3 for and 2 abstentions. The video and minutes of the meeting are available.

How do you intend to address the issue of existing Helium users with AU915 hardware already running on the network? Will these users be compensated if this change is approved?

I understand that AU915 is technically superior with additional channels, increased network capacity and coverage [1]. This seems to go against some of your more technical points. I'm no radio expert, but it's what I've been reading.

  • Four large public operators (NNNCo, Spark, Ventia) use AS923-1. The only large operator to use AU915 is TTN.

How is this relevant? Unless the Helium network is being built specifically for these operators to potentially roam on, instead of having decisions being made in the best interest of the Helium network in the long term.

  • Roaming with other countries in the region is easier as most use AS923-1 (eg: logistics tracking).

This decision shouldn't be based on what other countries use. There are several South American countries that use AU915 where logistics tracking could also be used.

  • Large organizations are requesting AS923-1 compatibility (Sydney Water, Urban Utilities)

Same with AU915 [2]

[1] https://www.thethingsnetwork.org/forum/t/future-strategies-for-au915-and-as923-in-australia/44218 [2] https://perth.wa.gov.au/en/live-and-work/smart-cities

Australian-Hnt-collectors commented 2 years ago

The thing the Dewi is over looking here is the following. AG TECH or the biggest possible network growth for the helium network is mostly AU915 which is why tenders From Government departments for various smart farm projects and smart wine or bio-solid treatments have requested 915 as a standard network use, as for their claim that NNNCO only use AU923 this is not the case and that is a down right misleading https://www.nnnco.com.au/wp-content/uploads/NNNCo_NSave_product-brochure.pdf also much of our energy sensors are AU915, also when you have companies like SIGFOX about to launch 915 radios for the Australian Market https://www.businesswire.com/news/home/20180417005723/en/Libelium-IoT-Sensor-Platform-Adds-LPWAN-Coverage-for-Australia-ASIA-PAC-and-LATAM-with-LoRaWAN-and-with-Sigfox-Technologies Maybe the DEWI LOWRAN team should actually consist of people whom come from Australia and are developing the network here for different use cases including low carbon and carbon neutral power companies, AG TECH and also people whom are currently in the market place rather then making a choice based on not entire truths as is the case with NNNCO as it uses both freqs in Australia. no offense to Anyone with this but understand please some companies have fought hard to get helium accepted indeed allowed to be used for data transfer for AG tech which is based on AU915 as accepted and when even the biggest telco company here Telstra offers a IOT solution on their own network but is planning to extend to au915. There is SO MUCH which has been missed and as for the 4 big networks in Australia currently there is 40 IOT network suppliers here so please rethink the 4 players https://www.iotone.com/supplier/searchlist?filterName=HQCountry&HQCountry=AU

leogaggl commented 2 years ago

Formal dissent

This PR seems to be riddled with inaccuracies and it is at best totally skewed towards one outcome (cherry-picking arguments that suit rather than a considered review of both options). As far as one can tell this whole process was driven by one company only and no other input was considered. Due to the fact that recordings have been lost and there was no transparency around this decision-making process, it is quite hard to come to any other conclusion.

I also note that the documents mentioned several times (Alper Yegin spoke about them in some details in the attached recording) have not been made available as part of this documentation. Can these documents please be shared for transparency? Especially since meeting notes attached to this PR are a bit of a joke. 3 paragraphs and a few bullet points? Seriously? That is not even remotely close to an accurate transcription of the meeting. And previous meeting notes are missing completely.

I will use this initial comment to just deal with the inaccuracies and omissions in this PR.

Inaccuracies

  • Four large public operators (NNNCo, Spark, Ventia) use AS923-1. The only large operator to use AU915 is TTN.

I can only count three operators. Only two of them in Australia. None of them is actually a public network operator. All of them are private. Also, what does large mean in this context? There is no data provided to make a considered opinion.

There are at least 4 network operators in Australia that use AU915.

There are several other smaller private networks in Australia also using AU915.

  • Roaming with other countries in the region is easier as most use AS923-1 (eg: logistics tracking).

These roaming arguments are a complete furphy. Are there any examples? All tracking device manufacturers are aware that there will NEVER be one frequency plan. Roaming devices will have to deal with country regulatory regimes and adapt. If you look at Australia's largest trading partners the majority is outside of any AS-923 (any of the 4 variants).

Australia's closest neighbour and BY FAR the largest population (Indonesia - population ~280 million) are on AS923-3 and PNG as well as NZ is on AU915.

Regions on Helium Network with AU915

According to the Helium Documentation, the following countries use AU915 (https://docs.helium.com/lorawan-on-helium/frequency-plans/#au915)

I have highlighted some of the larger ones. And it is worth noting that they share a lot of similarities with Australia in terms of their size and large regional use-cases. Again AS923-1 does not even come close to the area covered by AU915 or the population of these countries.

  • Large organizations are requesting AS923-1 compatibility (Sydney Water, Urban Utilities)

Says who? I note that there are no supporting documents for any of this. How many AU923 devices are said organisations using? There are a large number of Urban Utilities as well as Local Government Bodies using AU915 for current projects in Australia (SA Water and SA Power Networks are just two of them). There are also quite a number of public LoRaWAN tenders in Australia mandating AU915 as the official ACMA frequency band for Australia. Happy to provide evidence for this.

Omissions in favour of NOT changing to AU923

I am sure there will be plenty more discussion on the technical details. It is worth noting however that the actual two RF experts on the committee did abstain from the vote. Which is consistent with other members of the LoRa Alliance Technical Committee I have spoken to.

And just to be clear. I would welcome any company roaming onto Helium. But to downgrade over 4400 gateways (as of today and there will be many more by the time this comes to a vote) makes no sense. It is in fact highly dangerous and counterproductive.

EDIT: 2022-03-02 - added fourth AU915 LoRaWAN network operator EDIT: 2022-03-08 - added "formal dissent" EDIT: 2022-03-10 - edit "4 variants" as there is already an AS-923-4 apparently for Israel

Pavleau commented 2 years ago

This proposal appears to be at complete odds with the Helium motto of being the peoples network. Here we have a change request being lobbied for by one enterprise who doesn't even appear to have infrastructure in place in our country and it will be at the detriment to early adopters who have invested to establish a robust AU915 network. Come on Helium, please review this process and put all of the facts on the table.

Ravenitt commented 2 years ago

I wish to lodge a formal dissent.

Changing to As923 is going backwards. AU915:

Official ACMA Regional Standard 64 available channels (8 sub-bands) Higher antenna power (30 dBm max EIRP) no duty cycle limitation AS923:

16 channels (2 sub-bands) Limited antenna power (16 dBm max EIRP) Multiple implementations depending on which Asian country (AS923-1, AS923-2)

WilfridoT commented 2 years ago

Mmm and who's going to be responsable for all of the changes that we will have to do to get our devices and services running again ?

tonysmith55 commented 2 years ago

I'd like to review the three technical points "For AS923" in the proposal and invite a discussion. I will leave the importance (or otherwise) of regional roaming to others.

Downlink coverage on AS923-1 has 6dB high coverage thanks to using 125 KHz instead of 500 KHz (assuming the same spreading factor). That being said, gateways tend to have higher output compared to devices which allow for symmetrical link margin.

This is the not the issue of importance. The more important question is "how does an AS923 system perform in the presence of an Au915 network and vice versa"?

  1. How much capacity does an AS923 system loose in the presence of a Au915 network. And for that matter, how much capacity does it loose in the presence of the other non Lora system that also share the band.
  2. How much capacity does a Au915 system loose in the presence of AS923.

Capacity constraints affects both the end-user as well as the Network providers and in this case would affect and reduce the revenues Helium can obtain from the use of their backend infrastructure.

Let's start with some background information

AS923-1 is a dynamic channel plan which provides flexibility to easily expand/move uplink and downlink channels as the capacity increases or to deal with potential interference

This item is actually two quite distinct issues being put forward and they are:-

  1. AS923 allows for capacity expansion
  2. AS923 allows it to move to deal with potential interference

The LoraWan Regional Parameters (Pages 68 & 69) outline the use of the Frequency_Offset parameter to shift the AS923 channels up and down the band. Example on page 69 demonstrates how the channels can be moved down by 6.6Mhz to allow the use in countries which only have frequencies in the 915 – 921 band available. THIS MEANS, all the channels move down by 6.6MHz including the two common control channels. So I believe this bullet point by the DeWi’s LoRaWAN Working Group is misleading. It can be used to create a number of SEPARATE systems, each of 8 channels. For the user it would means at the time of commissioning a Node, they would need to select which of the potential separate systems it should be configured. To move a Node from one to another would require attending the site and reconfiguring the Node.
The only option would be to have a more sophisticated Node that can be reconfigured by over the air (gateway downlink) commands. This would be an inefficient and ineffective method to address capacity issues.

The same would apply to moving channels due to interference but again the solution requires people to make changes and would be slow to respond.

Compare this to AU915 where the system can scale from an initial 8 channels, to 16, 24, 32 etc all the way to a single system using all 64 channels uplink channels. The software in the nodes don't need changing, they are simply configured by the network at the time of a Join. Equally the network can reconfigure a node without a Join being required.

The Join Procedure is faster on 8-channel networks as unlike AU915, the AS923-1 Join channels are fixed. AU915 has a ⅛ chance of Joining on every attempt as it tries every band.

The concept of LoraWan is that a Node should Join a network and continue to transmit (uplink packets) for a very long time without the need to Rejoin. Every Rejoin is a significant drain on the network's capacity and so excessive number of Rejoins should be seen as coming from a problematic Node or a network implementation issue (eg a new gateway should be located closer to the node). Now if a Node Joins and continues to transmit uplink packets for months or even a year then the time to Join and the 1 in 8 chance of finding the network on the first attempt in an Au915 system is insignificant and the issue is completely irrelevant.

The issue is even made even less relevant as an Au915 network grows. Gateways expand to cover 16 rather than 8 channels and the the probability of joining on the first attempt halves to 1 in 4. So as a network grows an already irrelevant issue become even less important.

Tony

Bfindlay commented 2 years ago

Im sorry but this does not seem like a sound proposition. The peoples network should not be manipulated and changed due to the demands or needs of large corporations. Thats a no from me.

KarrathaKarl commented 2 years ago

I am also against the proposal to transition to rom AU915 to AS923-1. The guys commented prior have already clearly articulated the reasons and I share their concerns.

tonysmith55 commented 2 years ago

@lthiery, during the Zoom meeting I commented the Australian community had met a number of times to discuss AS923 vs Au915 to provide a recommendation to TTN. Looking back in my post on the TTN forum (where the written communication was undertaken) I can see this was in Feb 21.

In the early days of LoraWan in Australia there was a shortage of Au915 Nodes and a ready supply of AS923 Nodes. This came about as the Au915 band plan is not the same as US915 and so many suppliers had not developed suitable software. This is now a thing of history and the problem is long gone.

The important take-away message, AS923 in Australia was not related to inter-country roaming. To put the size of Australia into perspective. It's as wide as from Ireland to Moscow. Where the Europeans think of inter-country roaming, for us, this is simply moving around the one country. Just the Australian state of South Australia is about the same size as California plus Texas added together.

The recommendation proposed to TTN to incorporate legacy AS923 nodes was:

  1. Leave AS923 where it’s currently located as it will receive less interference and work better than moving it to AS-923_925 (PS, there was a discussion to move AS923-1 to AS923-2)
  2. Leave AU915 where it is in FSB2
  3. Continue to support existing AS923 systems - ie don’t remove existing gateways
  4. The Australian TTN community promotes all new nodes should be installed on AU915 (PS , ie people communicate this at every opportunity)
  5. If a AS923 node is replaced (failure) then commission its replacement on AU915.
  6. Gateways that are solely on AS923 are not a good idea.
  7. All Gateways should be on AU915 and if a second head is installed it can be on AS923.

In this same post there were other comments and feedback that are relevant and worth noting, for example, one response was:

I was always of the the opinion that AU915 was better from network perspective due both to the capacity issues you highlight (none overlapping channels).....

I asked if

Europe on EU868 where I would bet they envy both the number of channels we have and the separation of uplink from downlink frequencies.

and the response was

Absolutely envious yes!

Now the person providing these responses is an authority of Lora and LoraWan, he worked for Semtech directly involved with Lora technology and from day 1 of the acquisition of Cycleo (that's the company that originally developed Lora modulation)

The full discussion can be found at https://www.thethingsnetwork.org/forum/t/future-strategies-for-au915-and-as923-in-australia/44218/8

Aekraj commented 2 years ago

Keep Au915

rockinrobstar commented 2 years ago

In reading the LoRa Alliance RP002-1.0.3 LoRaWAN® Regional Parameters it defines the AU915 channel plan as "the regional parameters for Australia and all other countries whose band extends from 915 to 928 MHz spectrum". AS923 on the other hand is defined as "originally intended to apply to regions where the frequencies [915…928 MHz] are present in an unlicensed LPWAN band but MAY also apply to regions with available bands in frequencies up to 1.67 GHz."

Would a move to AS923 not be a violation of that first definition and thus make Helium not in compliance of the LoRa specification?

The second definition also gives the impression that the AS923 is a legacy or channel plan of last resort. It contains mechanisms for dealing with a variety of regulatory bodies requirements - all of which are unnecessary for Australia and just adds complexity for questionable gains.

Australia is the only country in Table 1 that has two 'X's in the last column. Other countries have a single 'X' with a footnote if other channel plans may apply. Is this an error in the table ?

tonysmith55 commented 2 years ago

@lthiery Having now watched the video of the Dec meeting of the Helium LoraWan Working Group, it struck me I didn't hear any discussion on pending developments to the Lora/LoraWan technology and how this would influence the recommendation between Au915 and AS923. Now I realise this video was just one of the group's meetings and so the topic may have been discussed previously.

While it's not in the PR notes, maybe you could provide some commentary on the implication of future developments.

Any network, be it a cellular, LoraWan etc can be considered to be in one of two modes. It starts out in Coverage Mode where a single gateway is placed on the highest point to achieve the maximum coverage. Some distant nodes have marginal coverage but the closer in units work fine. As the network grows and more gateways are added the reach to a distant node is no longer a problem and the network switches to Capacity Mode. The growth of Helium has been spectacularly successful and so many cities have already transitioned from "Coverage" to "Capacity" mode. With the network continuing to grow at very fast rates areas currently on the fridge will soon make this transition. As a result a significant proportion of the Helium network is already working in "Capacity" mode.

In parallel Semtech and the Alliance have been working on additional feature sets and technologies to increase the capacity of the systems. As a reference some of these include:

  1. Frequency Hopping Spread Spectrum (as compared to the current Chirp Spread Spectrum),
  2. Repeater operation and
  3. Node to node transmissions within a LoraWan umbrella.

Given a large proportion of the Helium network is already working in Capacity Mode a number of these (and other) developments will be fundamental to the future performance of the Helium network.

So my question: Which band plan (Au915 or AS923) performs the best after these developments are finalised and rolled out.

A supplementary question would be: If it has not been considered, why not and if not then could you quickly seek a response from the committee.

leogaggl commented 2 years ago

Keep Au915

@Aekraj - you might want to change your thumbs up to a thumbs down on the main PR above in that case.

TRIG01 commented 2 years ago

Massive no from me and it’s really a slap in the face to everyone building the network.

leogaggl commented 2 years ago

For those not willing or able to read all this text here is a one-sentence public service summary of what is going down here.

Reason for having this PR

A company bets on the wrong horse and is now lobbying behind closed doors to shoot the leading horse.

"Technical" summary

AS923-1 is the bike lane on the AU915 highway. (thanks @johnruciak)

EDIT: 2022-03-04 - fixed typo (thanks @jsloane)

Australian-Hnt-collectors commented 2 years ago

I think it is about time actually Someone from Dewi stepped forward and someone from Helium Management and became involved in this mess as already people are turning off units as they are over the corporate peoples network, and there is now four SME that i am aware whom are seeking advice as to the movement forward

matthewtropicpainting commented 2 years ago

If anyone changes the network then all of Australia will not support this anymore. Really think before anyone changes anything. Will be the end of helium in Australia. We do NOT want this

JaFaint commented 2 years ago

I do not support this change.

rockinrobstar commented 2 years ago

Here is a map I made just to compare for myself AU915 to AS923.1 (Apologies to some really small island nations that I might have missed or were too small for the map). I think it supports the notion that AU915 is appropriate to geographically larger countries located in South America and Oceania whereas AS923.1 seems to be more concentrated around South East Asian nations (and a few scattered others). Based on the data from here. AS923_1 vs AU915

Australian-Hnt-collectors commented 2 years ago

perth http://okosvaros.lechnerkozpont.hu/en/node/1052 Also do we forget that Perth was the first city to lock its smart citys into 915 freq as per the link and screen shot proves, this means that Helium could provide extra coverage and data transfer points if we stay on 915.

Australian-Hnt-collectors commented 2 years ago

perth 2 Also a point of note. this second image shows the lora alliance called for the city of Perth to use 915 freq and supported its roll out. I am pretty sure Perth would be asking a lot of questions if suddenly a freq change happens especially if they have sensors which cannot work on 923 or indeed where the cost is going to come from for access to a private network. this is really starting to look a tad fishy being honest.

Zodie666 commented 2 years ago

I wish to lodge a formal dissent to this change.

I am a helium hotspot owner and I DO NOT support the switch to AS923-1 due to the technical information supplied above.

tonysmith55 commented 2 years ago

The October meeting of the DeWi committee provides some interesting information about the technical capabilities of Helium backend and the issues they are already seeing in high density areas.

The meeting can be viewed at https://vimeo.com/681981386 with the first 15 minutes the most relevant to this HIP and PR.

The lead blockchain developer at Helium has proposed the person who is commissioning a Gateway can select a Sub-band for its operation. While not currently part of the Helium feature set, it is seen that this can be rolled out and added. The discussion also explained each sub-band, under the current HNT incentive scheme, would be seen as being on different networks.

What was interesting is that is would help solve a problem they are already seeing in the US with "Over Saturation". It was described it could assist in highly dense areas like San Francisco or New York to help open up another sub-band in addition to the existing Sub-band 2. It allows maintains the "ethos" of the Helium network of not being so much "Centrally Controlled".

The real benefit of this approach is that in an area with a lot of gateways where the rewards drop due to shear number and over saturation of gateways, there is an incentive to move to another sub-band. At the same time, in a sparse area, there is an incentive to have all gateways on the same sub-band to maximise the HNT rewards.

What was really interesting, is this affect is already happening in the US, there is already an over saturation of gateways in certain locations.

Now there are two aspects to Saturation:

We need to be careful in the way we think about congestion and traffic levels in an IoT system. In a Cellular system the measure is the number of people in a given area. eg the number of People per Square Kilometre. Even then we hear people thinking congestion only occurs in big cities, eg New York. It occurs everywhere, it just happens more often the larger the city. Congestion can occur in a cellular network due to Pedestrian Lights. People bank up, more and more people queue in a small area and when the lights turn green, a wave of people move down the street, all moving from one cellular base station to the next.

In the IoT world we need to be careful not to think of the number of people in a city but the number of water meters in a given area, the number of smart street lights and in an agricultural setting, for example, the number of cows with a tag. So the IoT world you can have a rural area with sparse coverage and within this a large concentration of cows on a dairy farm. The farm would have a number of gateways and even then at milking time, all the cows concentrate in one area which can easily overwhelm the network.

So in the IoT world we already need a network that can scale, a single gateway with sparse coverage scaling up to a number of gateways for a specific area all the way to multiple gateways with multiple sub-bands in specific locations.

Au915 already accommodates this, much the same as does US915. Even though some of the frequencies are different, the same overall network principles are common to both. All the proposals I've seen for AS923 are dependent on the creation of specialised and non-standard firmware in the nodes. eg AS923 nodes that don't obey the 14dBm limit but rise to 30dBm, nodes that move to a yet to be defined channel plan and so on.

So my question to the DeWi committee. Why did members vote for AS923 with its limitation in Australia while at the same time the meetings were discussing the BlockChain proposal and how this could assist with the Saturation already occurring in the US.

It just does not make sense. Why would Helium want to limit Capacity in Australia as this will affect their revenues.

tonysmith55 commented 2 years ago

Over at https://github.com/helium/HIP/issues/311 there is a parallel discussion on Au915 vs AS923. One of the proposed benefits of AS923 is " AS923 downlink will be able to reach further than on AU915 (SF12BW500). The difference is about 5 dB which is close to doubling the downlink range"

While I haven't analysed this in detail, I can expect this is probably true. Lower bandwidths (eg 125KHz vs 500KHz) transmit at slower bit rates and lower bit rates can be received more reliably. A parallel of this can be found in the days of using RS232 serial communication. A 1,200 bits/second RS232 transmission could go further and be more reliable than a 9,600 bits per second transmission.

However as we know, nothing ever comes for free. The downside is:

So while it sounds a solution, in reality I don't believe it's the answer

tonysmith55 commented 2 years ago

While there has been a lot of discussion around 8 channels of either Au915 or AS923, modern gateway designs are based on the SX1302 or SX1303 baseband chip which incorporates 16 receivers.

Lora modulation has the benefit that a gateway can successfully receive two nodes on the same radio frequency at the same time, provided they are using different spreading factors. They need to be what is called "Orthogonal" to achieve this.

Now the original gateways using the SX1301 baseband chip is programmed for 8 radio channels and has 8 receivers. They could receive 8 nodes on 8 different radio channels or other combinations including 2 nodes per radio channel across 4 different radio channels, provided each was using a different spreading factor.

The more recent SX1302 and SX1303 baseband chips incorporate 16 radio receivers and are programmed to listen on 8 radio channels. As a result gateways using these chips can receive 16 nodes at the same time, again provided any using the same radio channel are using different spreading factors.

Since all new gateway designs are based on the SX1302 or SX1303, I would be interested to hear what benefits AS923 has over Au915 with respect to these new Baseband receivers.

buzzware commented 2 years ago

@lthiery As per HIP45, could please confirm which of the above authors have met the requirement of proposing a "formal dissenting opinion with counter-arguments to the proposed change and must provide alternate solutions if possible". Assuming there is at least one, step 3 (a forum) is assured and those authors will need to be present and "must be ready to take the floor to represent their position" https://github.com/helium/HIP/blob/main/0045-lorawan-frequency-plan-selection.md

tonysmith55 commented 2 years ago

Now, i've heard and read comments that "particular Asian countries, with high populations and high population densities selected AS923 and countries like Australia with low population densities selected Au915".

These comments suggest the country regulator had a choice and made a technical decision on the frequency band based on their population.

This could not be any further from the truth, and in fact it's the other way around.

A regulator looks for bands of low utilisation and allocates or reallocates these for particular use, either on a primary or secondary basis. In the case of the Lora and LoraWan which use the unlicenced ISM (Industrial Scientific & Medical) bands these vary from region to region. In Europe the regulator was able to allocate frequencies in the 868Mhz region, in parts of Asia in the 923 Mhz part of the band, in China around 400MHz, the US a much larger block of channels was able to be allocated and they are in the range from 902-928Mhz. In the US there are enough frequencies in the ISM band for the Lora Alliance to specify 64 channels from the Node to Gateway AND a separate 8 channels from the Gateway to Node. The Australian regulator was not able to allocate as many frequencies, its restricted to 915-928 MHz and so the 8 downlink channels from the Gateway to Node overlap and the upper end of the 64 uplink channels used from the Node to Gateway. (As an aside, AS923 happens to reside in this overlapping region of the band and so it subjected to traffic from both Au915 nodes and Au915 gateways)

With the move to digital television in Australia the regulator was able to reclaim frequencies used by analogue television and this is an example of how frequencies are reclaimed and reallocated over time to other services. Dates were established per region for analogue television to shut down which provided certainty to the incoming services would have the necessary spectrum to operate.

A couple of examples of this associated with Lora and LoraWan.

Band clearing and reallocation takes years to achieve, but its an indication the limited spectrum in Europe is seen to be inadequate for the future needs of IoT type systems.

So, once again, I cannot understand why people on the DeWi Advisory Committee would vote to move Australia to a restricted set of frequencies when there is already a world-wide understanding more spectrum is required.

Australian-Hnt-collectors commented 2 years ago

@tonysmith55 100 percent agree with the above, it Also feels as though the Dewi are side stepping the Issues especially now they are aware of the Lora Alliance Helping and guiding Cities to use AU915 for the smart city roll out in Perth Also as with the TTN network which is 915 there are currently around Australia 44 communities using AU915 set up by TTN as per their press release last year alongside the smart cities initiative. Also with them Knowing now that we have federal backing for the AU915 to be rolled out on the helium network as named by a federal member of parliament for the smart mines project currently underway this Dewi lora committee made choices away from input from anyone on the ground here actually working in this space and it seems they do not acknowledge the weight of evidence for staying AU915. if you need to receive a copy of the letter of federal support ask Leo he has a copy Tony.

mroutsinaus commented 2 years ago

I’m in Buderim Queensland and currently have 1 hotspot. I do NOT support the switch to AS923-1.

buzzware commented 2 years ago

I wish to lodge a formal dissent to this proposal

Counter arguments

1) The framing of this proposal is wrong. Australia is not undecided on its frequency band, and there are not 2 equally appropriate future bands on offer for a network that has been operating on one of those bands for nearly a year. The burden of proof should be on the faceless commercial providers to prove to the much larger Australian "People's Network" community that they should abandon their existing commitments to AU915 and move to AS923 for some specific greater benefits. Instead they are using their privileged position on the Helium LoRaWAN committee to force a change on the community for their own benefit, and the committee has not been seen to have a genuine debate considering the interests of the existing Australian Helium community. 2) In the video meetings, most of the expressed reasons for a change come from the Actility representative. Actility has publicly announced partnerships with AS923 examples NNNCo and Spark, so the committee is not balanced or objective. In fact many of its members represent Helium's competitors. 3) While AS923 is used by private commercial operators in Australia, before Helium, TTN AU915 was the default public LoRaWAN service in Australia because the free tier was embraced by IoT user groups around Australia. As of November 2019 there were 43 TTN communities and over 500 gateways - the widest public coverage as you can see here https://www.thethingsnetwork.org/country/australia This is an easy market for Helium to convert because Helium is currently 100% compatible, only costs slightly more than free, supports commercial use, has dramatically greater coverage and an incentive for hosts. 4) Developers, startups and commercial contractors have fleets of devices, projects and deals in process (apparently even with Helium Inc) that assume or promise AU915 and can't be easily or cheaply changed, if at all. It is a great concern that the Helium LoRaWAN committee gives no regard for Australian users already committed to AU915. 5) The Australian TTN community already went through this same debate and arrived on AU915 as the best choice which can be seen here https://www.thethingsnetwork.org/forum/t/future-strategies-for-au915-and-as923-in-australia/44218 In particular, it offers greater antenna power (useful for Australia's vast distances and applications in farming and mining) and less contention. 6) Why did Helium choose AU915 in the first place? Likely for the same reasons that TTN chose it, and for a smooth transition from TTN. These reasons haven't changed. 7) Commercial operators are free to add AU915 to their offering if Helium roaming is so important to them. Helium is no hindrance to them or their customers from doing what they already were. Roaming is good for Helium Inc, Helium hosts, operators and their customers and the community would be more than happy for commercial operators to roam on Helium AU915. 8) The Australian Helium community have suffered enough. There was no manufacturer taking orders for Australia until February 2021 when there was around 20,000 hotspots. Then Nebra made promises and took orders and payments that were delivered 4-6 months later, when there were 4 to 7 times as many hotspots and that many times less earnings. There have been similar problems with Calchip, Linxdot and Syncrobit. Others have been very slow to or still have not supported AU915 even though it would only take software configuration and regulatory approval. So Australia was excluded from the huge HNT earnings that early US hosts enjoyed. We lost thousands of dollars due to the difference between promises and when they delivered. Things only got better when Sensecap appeared and IoTStore became distributor around September.

Alternatives 1) stay on AU915 like we always assumed that Helium would 2) Develop a scheme where a country may have a sequence of approved bands. If a hex already has active coverage for the first band in the sequence, allow or co-ordinate additional hotspots in the same hex to operate on subsequent bands in the sequence. This would ensure one band has the broadest coverage; and enable hosts to operate multiple hotspots in the same location on parallel bands with full earnings on each band. For Australia, AU915 would be the first band and not be interrupted. AS923 the second. Maybe the other 48 AU915 channels could follow.

shawaj commented 2 years ago

Just to weigh in here from Nebra.

We are also of the opinion that AU915 is the best option (and have been since this was first suggested here https://github.com/helium/miner/pull/991)

However, if there were to be a change to AS923-1 our hotspots would auto update to the new frequency plan immediately (this is already built in to our software).

Australian-Hnt-collectors commented 2 years ago

Just to weigh in here from Nebra.

We are also of the opinion that AU915 is the best option (and have been since this was first suggested here helium/miner#991)

However, if there were to be a change to AS923-1 our hotspots would auto update to the new frequency plan immediately (this is already built in to our software).

Good Afternoon Aaron, As your well aware the ability of the Nebra Unit is the reason we made the choice to deploy the 100 Nebras we have within the Sydney region. especially with the support we receive from nebra in the operation and deployment of the units, which is currently being showcased to the various departments within both the NSW AND Queensland government for AG tech and Mine site run outs for the Smart farms and Mines based on the AU915 freq in which we will be then looking to deploy the new Rockpi unit within regional NSW as i have discussed with your team. Hope you have a good rest of your week Aaron Many thanks Jayson

tonysmith55 commented 2 years ago

I'm seeking comments on some point being suggested about the capability of AS923. As an example I have heard comments like

As far as I can tell I can't see these in the Regional Parameters specification. For this I'm referencing the latest version being "Regional Parameter Version RP2-1.0.3"

Refer pages 68 & 69 of the "Parameters" document, I can see a standard system is specified to use the following Common Channels, which means each band of AS923 does not share common control channels.

Application Common Channels AS923-1 923.0 - 923.5 MHz AS923-2 921.4 - 922.0 MHz AS923-3 916.5 - 917.0 MHz AS923-4 917.3 - 917.5 MHz

I can only conclude we would end up with separate systems where a user would have to decide which network to configure their node to work with. To change this would require revisiting the site and reconfiguring the device. I could entertain this with a few devices but accessing 10,000 street lights is another thing.

Referring to page 67 Section 2.10.7 of the Parameters document. By Default, the Node listens in its receive window on the same channel as the preceding uplink.

From this I understand a standard AS923 system is simplex as compared to US915 and Au915 where the Node transmits an uplink on one channel and receives a response from the network (via a gateway) on a different channel. In doing this, downlink transmissions in a US915 or Au915 system do not consume capacity from that part of the RF spectrum used for uplinks.

I realise the Regional Parameters have other features but the terminology being used in previous discussions is "AS923-1", "AS923-2" etc which leads me to these parts of the document.

So, I would like to invite anyone to identify a particular feature in the Parameters document and explain how a standard public network would use this. Particularly how a distributed public network would manage and coordinate any special modifications.

leogaggl commented 2 years ago

@tonysmith55 it is indeed surprising that those pushing for such a major change which will have major ramifications for the future of Helium in Australia in some backroom have not in any way engaged in this "written discussion" as part of HIP-45.

@lthiery who were the people or organisations requesting this change (other than Actitility which was the obvious driver in the committee)? Where are those "Large organizations are requesting AS923-1 compatibility (Sydney Water, Urban Utilities)" voicing their support and reasoning?

We are still waiting for a copy of the document that underlies the vote in the committee. Can this please be made available as part of this "written discussion"?

All we have is the sparse information provided in this PR.

leogaggl commented 2 years ago

@lthiery

Formal dissent

I would formally like to request that Australia and New Zealand are to be treated as separate countries for the purpose of this HIP-45 process.

It is non-sensical to treat two countries with

Can this PR (28) please be changed from "Australia and New Zealand Switching to AS923-1" to "Australia Switching to AS923-1" and a separate PR be created for New Zealand.

EDIT: 2022-03-13 - added "formal dissent" - this will formally be a formal request at the town-hall meeting

lthiery commented 2 years ago

I'd like to make a few points about the process here.

  1. Everyone needs to stop making baseless allegations. If you don't have evidence, do not make public conjecture about another entity's motivations. Doing so is tantamount to libel and it drives a wedge in this community that unites local technologists, world-wide businesses, LoRaWAN enthusiasist and cryptocurrency enthusiasts. DeWi's Committee composed of members who are almost all on multiple LoRaWAN Alliance Technical Committees while also having regular roles at their respective companies. They volunteered on my request to provide technical guidance to the Helium Network on LoRaWAN topics and don't deserve the treatment given by many here. If you can't have a fact-based discussion about a technical decision being made, please disengage. I'm not going to call out the many comments in this thread that fall in this category, but I'd like to commend @tonysmith55 for his technical arguments and diplomatic tone. I believe @leogaggl is also working on improvements to HIP45 and the composition of the this LoRaWAN Committee and I commend such constructive efforts. This is a nascent network with evolving governance and it seems like many are happier to rail on how things are rather than roll up their sleeves and contribute.
  2. We are in day 9 of a minimum 28-day period for discussion and this whole process is resolved by a vote from hotspot operators concerned by the change. The initial publication of the reasoning for the change is far from the last word (in fact, it's the first word really) and is a summary of the arguments by the committee's members. HIP45 is meant to expedite "easy changes" while allowing for a drawn out discussion in the event of disagreement. Your disagreement is being heard and please be patient for a response from DeWi's LoRaWAN Committee. As mentioned above, this is one of the members' many roles and this committee is freshly formed and still trying to find the best process for everything.
leogaggl commented 2 years ago

@lthiery I don't think you guys in the US realise what this would mean for people affected here. And I (along with the rest of the community it seems by these comments here and in the Helium AU915 Discord as well as the relevant channels of the Helium Discord) can still not see any valid reasoning for such a massive disruption nor any long-term benefits. There has been a lot amount of reasoning provided by the community to the contrary. Thanks to @tonysmith55 for a large amount of detailed technical input and others for constructive comments.

I don't like people getting emotional about this either. But on the other hand the way this was communicated and against all the prior objections of the affected community (I am still finding Github comments with people objecting as early as July 2021), I can also understand the frustrations of local community members. The more I look into the history of this the worse this gets.

Example: https://github.com/helium/miner/pull/991

I stand by everything I have written here (and I am happy to correct myself if I have gotten anything wrong). Objectively on the evidence provided (one meeting recording, one meeting note and this PR) you can not conclude that this was driven by anything else than an extremely narrow interest of one company (and their 2 clients in AU and one in NZ). I have asked multiple times for the committee document to be shared as it is the only document referenced multiple times in the meeting recordings. Criticism of the way this decision was made is valid. And in the absence of any evidence (I have watched all the meeting recordings and gone through everything thoroughly), you can not come to any other conclusion. We have repeatedly asked for any evidence to the contrary, but none has been forthcoming.

We will keep within the process of HIP-45 and we are working hard on constructive improvements to the process. As per our community call, we are working on a HIP to try to address the issues with the committee process.

We would invite any assistance in this process. If there are any Open Source (or not-for-profit) governance experts here we would love their input. The very first DRAFT is in this HIP fork - please comment or raise a PR against this if you are able to assist.

lthiery commented 2 years ago

you can not conclude that this was driven by anything else than an extremely narrow interest of one company (and their 2 clients in AU and one in NZ)

I'm sorry, @leogaggl, but every civilized legal system abides by the presumption of innocence. There is no burden to disprove your baseless allegations. You have failed to substantiate anything beyond conjecture, so drop this accusation or step away from the conversation. "People being emotional" because of a proposed change does not entitle them to throw away basic decency.

Australian-Hnt-collectors commented 2 years ago

@lthiery could you please point out any comment in this thread which you think is Liable in your view, opinion, and mindset so i can have our legal team review them and if needs be ask the author to either modify or change their view, perception or opinion or correct the view or impression they have of the process to educate them

lthiery commented 2 years ago

@lthiery could you please point out any comment in this thread which you think is Liable in your view, opinion, and mindset so i can have our legal team review them and if needs be ask the author to either modify or change their view, perception or opinion or correct the view or impression they have of the process to educate them

We're not here to debate what's libelous or not. Nobody is threatening legal action either. All I'm asking is that people focus on facts instead of making up stories about corporate boogeymen. I'm amazed even this statement is controversial.

Australian-Hnt-collectors commented 2 years ago

@lthiery could you please point out any comment in this thread which you think is Liable in your view, opinion, and mindset so i can have our legal team review them and if needs be ask the author to either modify or change their view, perception or opinion or correct the view or impression they have of the process to educate them

We're not here to debate what's libelous or not. Nobody is threatening legal action either. All I'm asking is that people focus on facts instead of making up stories about corporate boogeymen instead of debating the factual merits of the proposal. I'm amazed even this statement is controversial.

I am Just trying to make sure we are on the same page, trying to keep this civil, polite and to the point without any comments or issues which may upset people as within Australia we are very mindful of peoples emotional states due to conjecture if there is any. Also mention Liable to an Australian or Australian community and they will consider if they have been or not as the last thing anyone wants is potential legal action

lthiery commented 2 years ago

@Australian-Hnt-collectors https://github.com/dewi-alliance/hplans/pull/28#issuecomment-1063470809 If you need a concrete example, I explain myself two comments above. We all have limited time and there is no burden to deny a claim that itself is without evidence. I wish I had time to address some of the other great points & questions here but instead I find myself repeatedly asking people to simply be curteous & factual.

leogaggl commented 2 years ago

@lthiery I have no idea where all this libel talk is coming from.

I think the question of why this "proposed change" keeps coming back despite the lack of support and having been reversed once before is very legitimate. All we are trying to understand is why and where this is driven from? It certainly is not from the local community. It doesn't seem to have support from gateway manufacturers. It does not seem to stack up on technical merit (and we have provided ample evidence to that I believe). And from listening to the meeting recording, reading notes and PR this seems to be the only conclusion.

There are no comments from anyone in support of this change and no documentation underlying this decision is being provided. This is a question from watching the recording and not an 'allegation'. If it has come across as an 'allegation' - it is certainly not intended as such.

Otherwise, I have no intention of dragging this out further or adding more noise. I hope some answers to questions raised and the documentation will be provided at some stage in this process.

Australian-Hnt-collectors commented 2 years ago

@lthiery With all due respect to you i believe that people have been polite courteous and are trying to be factual based on the information available even when sometimes not all the information has been provided, in fact if i am honest this is the most civil i have seen many Australians whom stand to loose money on investments into the network, some of those losses could potentially total 250k to 500k if not more they have placed into the helium network and the support thereof. I trust the community will continue to be this respectful as we move down this road. Have a good rest of your day dealing with addressing the issues and points above wish you well in doing so.

lthiery commented 2 years ago

@leogaggl

You write two comments back:

you can not conclude that this was driven by anything else than an extremely narrow interest of one company (and their 2 clients in AU and one in NZ)

And yet in your last comment:

If it has come across as an 'allegation' - it is certainly not intended as such.

And in the same comment:

I have no intention of dragging this out further or adding more noise.

You clearly make an allegation, deny making an allegation and discuss it, and then say you don't want to drag it out and discuss it. I can only ask you so many times to please stop.

jaytheblader commented 2 years ago

Having been involved in this argument from the very beginning and having been a primary objector to the initial attempt to change the frequency in Australia I do find it odd that I was not invited to the DeWi meeting for my opinion. Going forward though I can see that Tony has provided more than 10 times the necessary technical information to clear up the nonsense of AS923 being superior over AU915 and the Australian community has overwhelmingly voiced their opinion on a preference going forward wouldn’t it be prudent for DeWi to formally reverse the proposal of changing Australia to AU915 lest we waste more time discussing something that it seems very few want.

buzzware commented 2 years ago

@lthiery In the December 10 video, only 2 people present provide reasoning regarding the change.

Alper (Actility) provides technical arguments that are outweighed by Tony Smiths comments. He also talks about roaming. Actility have announced partnerships with at least 2 of the 3 existing AS923 operators mentioned. Actility & NNNCo https://www.actility.com/nnnco-and-actility-ready-to-begin-rollout-of-australias-first-nationwide-industrial-iot-network/ Actility & Spark https://www.actility.com/lorawan-connected-boatsecure-a-sailors-delight/

Dave (Senet) is very ambivalent about the change and abstains.

Then there is a long silence, when anyone could have offered more points.

At no point was the community investment in AU915 by probably thousands of hosts mentioned.

So its just a fact that there is one network provider pushing this, with voting support from Semtech and Kerlink.

Thats why people are upset and pointing at one commercial operator, because only one is pushing for it.

An objective observer would wonder why "fix" what isn't broken, and why the push for the change has been so persistent with such weak arguments, and why there was no attempt to "sell" the benefits to the community as greater than the costs of change. This scenario breeds distrust.

lthiery commented 2 years ago

Having been involved in this argument from the very beginning and having been a primary objector to the initial attempt to change the frequency in Australia I do find it odd that I was not invited to the DeWi meeting for my opinion. Going forward though I can see that Tony has provided more than 10 times the necessary technical information to clear up the nonsense of AS923 being superior over AU915 and the Australian community has overwhelmingly voiced their opinion on a preference going forward wouldn’t it be prudent for DeWi to formally reverse the proposal of changing Australia to AU915 lest we waste more time discussing something that it seems very few want.

@jaytheblader HIP45 was developed to arbitrate these disagreements following the initial objections in September. There was not a single complaint about the process proposed.

We are engaging the process and Tony has done a fantastic job of making detailed arguments against the change. But you seem very upset with the process nonetheless, so I encourage you to draft a proposal for improving governance. However, this is not the place for it. Please see https://github.com/helium/HIP